1 / 29

Search for SUSY with early LHC data

Tommaso Lari Università and INFN Milano. Search for SUSY with early LHC data. Introduction how SUSY events look like at the LHC SUSY searches ( review of analysis strategy and discovery potential ) Search strategy and reach in parameter space

deiondre
Download Presentation

Search for SUSY with early LHC data

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Tommaso Lari Università and INFN Milano Search for SUSY with early LHC data • Introduction • how SUSY events look like at the LHC • SUSY searches (review of analysis strategy and discovery potential) • Search strategy and reach in parameter space • SUSY searches (realistic background estimation, roadmap to discovery) • Improved background computation: Matrix Element vs Parton Shower • Measurement of background from data: top, W+jets, Z+jets • Detector effects: ETMiss tails in QCD events, fake leptons • The first mass measurements • An other scenario: GMSB models

  2. Supersymmetry For every Standard Model fermion (boson) a bosonic (fermionic) partner is introduced. Solves problems with Higgs mass naturalness if the new particles are lighter than about 1000 GeV. At least two Higgs doublets needed: 5 Higgs bosons (+ 5 Higgsino fermions) New particles: Spin 0: scalar quarks, leptons, neutrinos Spin ½:4 neutralinos (mix of Zino, photino, Higgsino), 4 charginos (mix of Wino and Higgsino), gluino Mass terms for the new particles can be added to the Lagrangian (SUSY breaking) In general, 105new masses, mixing angles, CPV phases

  3. Minimal SUGRA • A random choice of the 105 MSSM parameters violates limits from B/D/K physics, electric dipole moments, flavour-violating neutral currents, … • Need some assumption on the structure of SUSY breaking lagrangian. In mSUGRA (5 free parameters, most studied by ATLAS and CMS): • Conserved R-parity (+1 for SM, -1 for SUSY particles): SUSY particles are produced in pairs, lightest susy particle is stable (good candidate for Dark Matter) • Common mass m0 for susy scalars, m1/2 for fermions (at GUT scale). • May be too constrained. Experiments are mostly interested in identify signatures to develop and study search strategies.

  4. p p ~  ~ ~ ~ q ~  l g q q l l • In SUGRA models strongly interacting sparticles (squarks, gluinos) dominate production. • Cascade decays to the stable, weakly interacting lightest neutralino follows. • Event topology: • high pT jets (from squark/gluino decay) • Large ETmiss signature (from LSP) • High pT leptons, b-jets, -jets (depending on model parameters) • R-Parity Violating models: the LSP decays, more jets and less missing energy. SUSY events topology A typical decay chain: An expample of mass spectrum:

  5. The standard discovery plot Most general search strategy: jets + ETmiss + n-leptons SM (PYTHIA) Meff =|pTi| + ETmiss (GeV) ATLAS Physics TDR Jets + ETmiss + 0 lept. • Backgrounds: - Real missing energy from SM processes with hard neutrino tt, W+jets, Z+jets • Fake missing energy from detector Jet energy resolution (expecially non-gaussian tails) critical A good understanding of both SM physics and detector (missing energy expecially) critical to claim excess over SM predictions 10 fb-1 1 TeV SUSY 10 fb-1 ATLAS SUSY selection cuts used in the pictures: • 1 jet with pT >100 GeV, 4 jets with pT>50 GeV • ETMISS > max(100 GeV,0.2Meff) • Transverse sfericity ST>0.2 • No isolated muon or electron with pT>20 GeV

  6. LHC discovery potential for Supersymmetrywell documented since several years ATLAS: Physics TDR, CERN/LHCC 99-14 CMS: J. Phys. G28 (2002) 469. • 1 fb-1 of data already allows discovery if squark or gluino mass < 1.5 TeV (as it should, because of naturalness). ATLAS and CMS potentials similar • Those studies assumed a perfectly known SM physics (only stat. errors on background rate) and ideal detector (nominal asymptotic performance). • SUSY discovery likely to depend not on statisticbut on the understanding of SM physics background and detector systematic with early data. • In this talk, emphasis on these issues M1/2 (GeV) g(2TeV) 1 fb-1 No EWSB M0 (GeV) LHC discovery reach • A scan of mSUGRA parameter space is performed and the S/√B significanceis evaluated for several signatures. 100 fb-1 2.5 TeV q q(2000) q(1TeV) 1 TeV g q(1000) g(1TeV) 1 TeV q LEP excluded

  7. Reach of different channels Inclusive ETMiss + jet: • Best signature • Important for high ∫L limit Multi-lepton n(≥1) ℓ: • Less powerful • But may be very useful for early discovery: • Signal confirmed in several channels • Better S/B, leptons better measured/understood than jets at the beginning – can be important in early searches Esempio: 2SS ATLAS

  8. Preparing for real data Most traditional studies uses an ideal detector (no miscalibration, misalignments, etc.) and SM background is simulated with parton-shower montecarlos and assumed to be “known” (significance is S/√S+B ) More realistic studies: Montecarlo production with Parton shower + Matrix element Take into account uncertainties on background cross section, rely on data as much as possible to evaluate SM contribution Emphasis on Full Simulation studies Increasing realism of detector effects

  9. Matrix elements and backgrounds tt+jet • Hard jet emission in SM processes is important for SUSY searches background • Standard analysis use Parton Shower Montecarlo for SM processes: badly underestimates hard jet emission. • Recent ATLAS background studies: • Generate hard process with exact ME computation (Alpgen, Sherpa, …) • Parton shower hadronization with HERWIG, PYTHIA • Solve double-counting problems with MLM matching PT of additional jet (GeV) Parton shower is a good model in collinear region, but fails to describe hard jet emission

  10. Backgrounds with MEs (1) SM (PYTHIA) SM (ALPGEN+PYTHIA) Meff (GeV) • Background increases But discovery of 1 TeV SUSYstill easy with only statistical errors Warning: susy signal is not exactly the same in the two plots (same mSUGRA point, but left plot use ISAJET to compute particle masses, right plot uses PYTHIA) ATLAS Phys. TDR (0 leptons) ATLAS Preliminary (0 leptons) SUSY (1 TeV) SUSY (mass=1 TeV) Effective mass (GeV)

  11. Backgrounds with MEs (2) • Additional jet, but not missing energy in background process: importance of missing energy crucial ATLAS Preliminary (0 leptons) ATLAS Preliminary (0 leptons)

  12. Backgrounds with MEs (3) SM (ALPGEN+PYTHIA) 1-lepton channel more promising than 0-lepton - Background decreases more than signal • Dominant background is top, more controllable than QCD jets (see later) ATLAS Preliminary (1 lepton) SUSY (mass=1 TeV) Effective mass (GeV)

  13. Background from data • Z+jets: big contribution from Z →  • Can use Z →ee, apply same cuts as analysis, substitute ET(ee) with ETmiss and rescale by BRs. • Top mass reasonably uncorrelated with ETMISS • Select events with m(lj) in top window (with W mass constraint – no b-tag used). Estimate combinatorial background with sideband subtraction. • Normalize to low ETMiss region (where SUSY small) • Procedure gives estimate consistent with top distribution also when SUSY is present ATLAS Preliminary Full Simulation0.5 fb-1 Blue: tt(MC@NLO) Green: SUSY Dots: top estimate

  14. Background from detector Jet resolution not critical for SUSY searches, non-gaussian tails are much more critical Can map badly behaving cells (simmetry, Z+jets balance, …) Avoid problemating regions (events with jet in crack) Veto events with ETmiss vector along a jet Modeling of detector response in simulation Problem: jet cross section very large, problems from a very small fraction of events in tails – difficult to produce the required statistic in full simulation Light Jets misidentified as leptons can contribute to 1-lepton channel background Lepton efficiency less critical But reduces significance of 1-lepton channel if low Also 3-lepton channel may be promising for discovery Must be understood for reconstruction of specific decays (see exclusive analysis later)

  15. SUSY mass spectroscopy p p ~  ~ ~ ~ q ~  l g q q l l mSUGRA • After discovery: reconstruction of SUSY masses. Two undetected LSP: no mass from one specific decay. Measure mass combinations from kinematics endpoints/thresholds. With long enough decay chain, enough relations to get all masses. • A point in parameter space is chosen, and decay chains are reconstructed. • Analysis should be applicable whenever the specific decay do exist. • Leptonic (e/) decay of χ02 “golden channel” to start reconstruction. Can also be a good channel for discovery • Some benchmark points favoured by cosmology studied in detail with full simulation

  16. Dilepton Edge Clear signature, easy to trigger: starting point of many mass reconstruction analyses. Can perform SM & SUSY background subtraction using OF distribution e+e- + +- - e+- - +e- Position of edge (LHC Point 5) measured with precision ~ 0.5% (30 fb-1). ~ ~  ~  l l l Polesello et al., 1997 e+e- + +- - e+- - +e- e+e- + +- 5 fb-1 FULL SIM Point 5 ATLAS ATLAS 30 fb-1 atlfast Modified Point 5 (tan() = 6) Physics TDR SUSY backg Mll (GeV) Mll (GeV) SM backg (invisible…)

  17. p p ~  ~ ~ ~ q ~  l g q q l l lq edge llq edge 1% error (100 fb1) 1% error (100 fb-1) Formulas in Allanach et al., hep-ph/0007009 Mass reconstruction: a typical decay chain The invariant mass of each combination has a minimum or a maximum which provides one constraint on the masses of l q ~ ~ ~ ~ ATLAS Fast simulation,LHCC Point 5 ATLAS TDR ATLAS TDR ATLAS TDR ATLAS TDR llq threshold ll edge

  18. Full simulation: 2-lepton edge p p ~  ~ Edge at (99.8 ± 1.2) GeV ~ ~ q ~  l g q q l l e+mu 1 fb-1 ATLAS Preliminary Point 5a - 4.37 fb-1 No SM Standard Model and SUSY combinatorial Backgrounds can be subtractedusing The combination: (e+e-) + β2(η) (μ+μ-) – β(η) (e+μ-) CMS Point LM1 is the acceptance correction factor for different e/ efficiencies (from Z decays?) The edge in dilepton invariant mass is a clear signature, with a very high S/B ratio tt bcg. only

  19. Jet+lepton combinations p p ~  ~ ~ ~ q ~  l g q q l l Several edges (mass relations between SUSY particles) may be visible with only a few fb-1 of data. Good jet resolution (and missing energy used in cuts) more critical Larger of M(llq) Smaller of M(llq) ATLAS Preliminary 4.20 fb-1 501 GeV 272 GeV

  20. GMSB scenario τ1is NLSP N1 is NLSP cτ>>L Like an heavy μ Like mSUGRA cτ~L NLSP decays in the detector, possible lifetime measurements cτ<<L Decay into 2τ Decay into 2γ L=detector size • τ trigger and reconstructionin early data not trivial • Decay into 2γ promising (good ECAL performance early enough?) • Lifetime measurements: need to understand vertexing in early data • For longer lifetimes, need to understand background: • Hard radiation from high-pT cosmic muons • Delayed hadronic showers (K0L and neutrons) In gauge mediated supersymmetry breaking models, the lightest SUSY particle is the gravitino. Phenomenology depends on nature and lifetime of the second lightest state: ~ ~

  21. GMSB Heavy slow “stable” leptons can be tagged with Time-Of-Flight measurements in muon drift tubes. Timing/triggering issues most critical? CMS Also similar ATLAS studies (Phys. TDR)

  22. Conclusions • Supersymmetry is one of the most promising extensions of the Standard Model. • In most models, a few fb-1 of data will allow the LHC experiments to measure a clear excess over the SM contribution and reconstruct several mass relations. Whether we can achieve this within the first year of physics run will depend on the ability of the experiments to understand their detector and the SM processes in a short time. • Recent ATLAS and CMS studies focus on • Understanding of SM backgrounds with the use of the latest Montecarlo tools, and development of strategies to validate the MC predictions with data. • Large scale productions of full simulation data, are used to study detector systematic and prepare for real data analysis. • Looking eagerly forward to the first data!

  23. Backup slides

  24. ATLAS Full simulation studies Focus Point LEP excluded • Goals: test software for data reconstruction and analysis, computing grid production. Study detector-related systematic. Validate fast simulation results. • 10M events produced in ATLAS Data challenge of 2005. • Five mSUGRA models studied. Focus on cosmologically interesting regions. • Typical statistic of 10 fb-1 • Some 10M events to be produced in full simulation in first half of 2005 with more realism (misaligned detector, calibrations, …) – focus on first 100 pb-1 of data Coannihilation SU1 SU2 Bulk SU3 No EWSB

  25. ATLAS Detector studies • Detailed studies of detector effects ongoing in ATLAS. • Jet energy response, lepton efficiency and fake rates studied with full simulation can be implemented in fast simulation to get high statistic samples with realistic detector response

  26. SUSY Higgs sector 2 doublets, 5 physical states: h0,H0,A0,H (mix if CPV) h light, SM-like. m  133 GeV Lots of free parameters in MSSM Often assume heavy SUSY states (no Higgs decay into SUSY nor Higgs production in SUSY decays) Define banchmark scenarios. Example:Carena et al. , Eur.Phys.J.C26,601 MASSH– maximum h mass allowed by theory Nomixing– small h mass (difficult for LHC) gluophobic – reduces hg coupling (and LHC production xSection) Small reduces hbb and hcouplings (harms some discovery channels)

  27. SM Discovery Potential: New and Updated Analysis 2004 2001 ATLAS 30 fb-1 ATLAS 30 fb-1 new : Vector Boson Fusion (VBF), H and H WW affects discovery potential for light Higgs (relevant for the h boson in SUSY) Issues for early data taking: tau reconstruction, forward jet reconstruction and trigger, veto on central jets (underlying event, …)

  28. SUSY Higgs scans ATLAS ATLAS ATLAS 300 fb-1 ATLAS 300 fb-1 h discovery curves H/A discovery curves LEP limit depends on top mass (here mtop = 175 GeV). No tanlimit for mt > 183 GeV Statistic is 30 fb-1 or 300 fb-1 depending on channels. Stat. errors only. Always at least one Higgs is seen (also for the other scenarios). Over a large parameter space, only h is observable and discrimination from SM Higgs is very difficult.

  29. Other SUSY-Higgs studies Higgs in cascade decays. Peak in bb invariant mass distribution – with SUSY cuts may be much easier to see than SM Higgs. ATLAS 300 fb-1 M1/2 (GeV) CMS M1/2 CMS  CMS Discovery with 10 (100) fb-1 Mbb (GeV) bb m0  M0 (GeV)

More Related